Production of prostaglandin E2 by tumor cells in vitro.

Our previous observations on the production of prostaglandin E2 (PGE2) by bladder tumor cell lines in vitro and the enhancement of tumor cell PGE2 production upon exposure to purified peripheral blood lymphocytes from normal human donors prompted us to examine this interaction in an animal model in order to further define conditions that determine the occurrence of this phenomenon. Cell lines derived from carcinogen-induced bladder and mammary tumors and from embryo fibroblasts in Fischer rats were exposed to purified peripheral blood or splenic lymphocytes in the presence or absence of indomethacin (10(-7)M). After varying times at 37 degrees, supernatants were harvested for determination of PGE2 by radioimmunoassay. Time course studies demonstrated rapid PGE2 production with plateau levels appearing at 8 hr. Increased tumor cell PGE2 production occurred in the presence of increased numbers of lymphocytes. Indomethacin partially inhibited PGE2 production. Preincubation studies suggested that the contribution of lymphocytes to overall PGE2 production in the present system was minimal. On the basis of previous observations of PGE2-associated inhibition of lymphocyte cytotoxicity against tumor cells in vitro, the present results suggest that tumor cell PGE2 production may reflect a response of the tumor cells to challenge by effector lymphocytes and may represent a mechanism whereby tumor cells subvert an immune response mounted against them.

Evidence that the antiproliferative effect of verapamil on afferent and efferent immune responses is independent of calcium channel inhibition.

Calcium channel blockers are capable of inhibiting the afferent and efferent limbs of the immune responses of human peripheral blood mononuclear cells in in vitro systems. This effect is thought to be related to the ability of the calcium channel blocker to limit the transmembrane flux of calcium. We report herein that two optical enantiomers of verapamil, one (S-) which is capable of blocking the slow calcium channel and mitogen-stimulated 45Ca++ uptake into human lymphocytes, while the other (R+) is incapable of either activity, share almost identical capabilities of depressing both the afferent and efferent limbs of immunity. These observations suggest that the inhibitory effects of verapamil on various afferent and efferent immune events are, in part at least, unrelated to the inhibition of transmembrane calcium flux.

Additive inhibition of afferent and efferent immunological responses of human peripheral blood mononuclear cells by verapamil and cyclosporine.

We have studied the effects of verapamil (0-50 microM) on the in vitro immunological function of human peripheral blood mononuclear cells in the presence or absence of cyclosporine (0-600 ng/ml). The proliferative response to phytohemagglutinin, OKT3, and alloantigens, the generation of cytotoxic T lymphocytes following allogeneic stimulation, and mitogen-induced reduction of intracellular ATP were inhibited in a concentration-dependent fashion by verapamil alone and by cyclosporine alone. When the two drugs were added to the same culture, additive inhibition was observed. A verapamil concentration of 5 microM usually reduced by at least 50% the amount of cyclosporine necessary to cause the same level of inhibition seen when no verapamil was present. The additive inhibition of the two drugs was likely not due to additive inhibition of IL-2 responsiveness, since neither drug alone inhibited the response of an IL-2-dependent T cell clone (CTLL-2) to recombinant IL-2 except at the highest concentrations tested, where a mild additive effect was noted. Nor was the additive inhibition related to an additive effect on total IL-2 receptor expression since an additive inhibitory effect on PHA-induced IL-2 receptor expression was only seen with 50 microM verapamil, while additive functional effects on mitogen- and antigen-induced proliferation and alloantigen-induced CTL generation were seen with 5 microM verapamil doses. Verapamil or cyclosporine alone inhibited IL-2 production of PHA- and phorbol ester-stimulated peripheral blood mononuclear cells--however, no additive effect was seen when the two drugs were both added to culture, probably because of the very potent inhibition by cyclosporine alone. Natural killer cell activity of human peripheral blood mononuclear cells against K562 target cells was significantly inhibited by verapamil in a concentration-dependent fashion, while cyclosporine had a more modest concentration-dependent effect. The combination of both drugs demonstrated additive inhibition. Effector function of cytotoxic T lymphocytes was modestly inhibited by either verapamil or cyclosporine alone. A combination of the highest concentrations of verapamil and cyclosporine caused an additive inhibitory effect. In summary, these data demonstrate that verapamil and cyclosporine have concentration-dependent inhibitory activities on both the afferent and efferent limbs of immunity that were additive when verapamil was used in a concentration of at least 5 microM. The additive effects are probably not related to effects on IL-2 circuitry.

The immunosuppressive properties of enisoprost and a 5-lipoxygenase inhibitor (SC-45662).

Metabolic derivatives of arachidonic acid such as prostaglandins and leukotrienes may alter immune responses. Enisoprost (ENO), a synthetic prostaglandin E analog, inhibited the response of human peripheral blood mononuclear cells to phytohemagglutinin in a concentration-dependent fashion: .1, 1.0, and 10 micrograms/ml yielding 4.0, 21.7, and 74.3% inhibition, respectively. ENO also potently inhibited both IL-2 production (measured by ELISA) and IL-2 responsiveness (measured by CTLL-2 response to IL-2) in a concentration-dependent manner, yet did not inhibit acquisition of IL-2 receptors. ENO similarly diminished efferent immune function in a concentration-dependent fashion, as measured by inhibition of cytotoxic T cell and natural killer effector function. A 5-lipoxygenase inhibitor (5-LO), SC-45662, also inhibited mononuclear cell response to PHA in a concentration-dependent manner: 0.1, 1.0, and 10 micrograms/ml yielding 5.0, 9.7, and 79.7% inhibition, respectively. Although 5-LO potently inhibited IL-2 production, it had no effect on IL-2 responsiveness or IL-2 receptor acquisition. Like ENO, 5-LO impaired CTL and NK effector function yet was not as potent in inhibiting CTL effector function. In summary, ENO and 5-LO inhibit afferent and efferent immune function. The inhibitory effects of these drugs are not related to cytotoxicity as cell viability is maintained for 72 hr in the presence of these drugs, and the inhibitory effect is reversible when the drugs are removed. The 5-LO does not inhibit mononuclear cell responses simply by shunting the formation of arachidonic acid precursors to form inhibitory prostaglandins, since it does not impair IL-2 responsiveness in a manner similar to ENO. These two compounds may prove to have clinical utility in organ transplantation if safely achieved serum concentrations of these drugs yield in vivo immunosuppression parallel to our in vitro results.

Regulation of the epithelial cell-specific integrin, CD103, by human CD8+ cytolytic T lymphocytes.

BACKGROUND: The destruction of the graft epithelium by CD8+ cytolytic T lymphocytes (CTL) is an important aspect of organ allograft rejection. Our recent finding in a mouse model that the epithelial cell-specific integrin, CD103, defines a subset of CD8+ CTL potentially sheds new light onto such interactions. The goal of the present study was to assess the relevance of these data to the human system. METHODS: CD103 expression by human T-cell populations generated in mixed lymphocyte cultures or isolated from transplant nephrectomy specimens was quantitated using multiparameter FACS analyses. RESULTS: CD103 defined a major subset (26-76%) of CD8+ CTL generated in human mixed lymphocyte cultures; cell sorting experiments confirmed that the CD103+ and CD103- subsets both possess allospecific lytic activity. Anti-transforming growth factor (TGF)-beta blocked the appearance of the CD103+ CTL subset, and persistent expression of CD103 by CD8+ CTL was dependent on bioactive TGF-beta. Isolated CD103+ and CD103- CD8 subsets maintained their phenotypic integrity during in vitro expansion, although optimal CD103 expression on the former was TGF-beta dependent. Although CD103+ cells were rare among activated CD8 cells in peripheral lymphoid compartments (< 10%), analyses of transplant nephrectomy specimens revealed that a major subset (21-61%) of CD8 memory/effector cells that infiltrate rejecting renal allografts express high levels of CD103. CONCLUSIONS: We conclude that CD103 defines a discrete and stable subset of human CD8+ CTL and that CD103 expression by such cells is initiated and maintained by bioactive TGF-beta. These data point to the existence of a human effector subset that is uniquely specialized for the destruction of the graft epithelium.

Enhancement of natural cytotoxicity in lymphocytes from animals with carcinogen-induced bladder cancer.

Splenic lymphocytes from Fischer rats with carcinogen (FANFT)-induced bladder cancer had depressed natural cytotoxicity that could be enhanced in vitro by the addition of mouse leukocyte interferon to the cytotoxicity assay. Such enhancement appeared to reflect an effect directly on lymphocytes rather than a cytotoxic effect on tumor target cells. The possibility that tumor cell prostaglandin production might partially inhibit lymphocyte cytotoxicity and its enhancement prompted separate attempts to enhance cytotoxicity by inhibiting prostaglandin production during cytotoxicity testing. However, addition of indomethacin to the cytotoxicity assays did not enhance cytotoxicity in lymphocytes from either control or tumor-bearing rats and did not add to the enhancement seen with interferon. Addition to the cytotoxicity assay of unstimulated peritoneal monocytes which themselves have been shown to produce prostaglandins, did not effect lymphocyte cytotoxicity. Correspondingly, indomethacin in parallel samples did not alter baseline levels of cytotoxicity seen. Further stimulation of cytotoxicity by addition of interferon to these samples was also not seen. Taken together, in vitro enhancement of depressed lymphocyte cytotoxicity in tumor-bearing animals was possible with exogenous leukocyte interferon, could not be accomplished by inhibition of prostaglandin production in this system, and did not appear to be influenced by the addition or deletion of monocytes during cytotoxicity testing.

Indomethacin and poly I:C in the inhibition of carcinogen-induced bladder cancer in an experimental animal model.

Fischer rats, in whom superficial transitional cell cancers of the urinary bladder were induced by the carcinogen N-[4-(5- nitrofuryl )-2-thiazolyl] formamide, were inoculated intraperitoneally with either phosphate buffered saline, indomethacin (a prostaglandin synthetase inhibitor), poly I:C (an interferon inducer), or indomethacin together with poly I:C. While indomethacin alone appeared to have a significant albeit variable inhibitory effect on tumor size, poly I:C had a far more pronounced significant inhibitory effect. The combination of poly I:C and indomethacin together, however, led to the greatest inhibition in tumor growth, and in some instances, to tumor regression. Splenic lymphocytes from the same animals demonstrated enhanced natural cytotoxicity after treatment with poly I:C. Surprisingly, levels of natural cytotoxicity seen in animals treated with indomethacin and poly I:C together were lower than those seen with poly I:C alone. No enhancement of cytotoxicity could be demonstrated in vitro in lymphocytes from indomethacin-treated animals. Macrophages were also treated in this system under identical conditions. However, the activity of macrophages alone and of macrophages and lymphocytes together did not appear to be modified either by indomethacin alone or by the combination of prostaglandin synthetase inhibition and interferon induction together, the combination of which in vivo was suggested to be most effective in controlling tumor progression. Further studies to determine timing of these interactions and doses of immune response modifiers in order to characterize mechanisms possibly at work in modifying tumor growth in this system therefore seem highly indicated.

Macrophage depletion and manipulation of enhanced immune response in an animal model of bladder cancer.

In a transplantable model of bladder cancer the inhibition of tumor growth by polyinosinic:polycytidylic acid (poly(I:C], concomitant stimulation of natural killer cell activity, and inhibition of both effects by prior inoculation with anti-interferon antiserum have been previously observed. The possible role of macrophages in these interactions is explored. Depletion of peritoneal macrophages by silica at concentrations that did not markedly depress natural killer cell activity was found to inhibit the enhancement of natural killer cell activity by subsequent in vivo poly(I:C) inoculation. Inoculation with silica also seemed to abrogate the tumor inhibitory effect of poly(I:C) treatment. To determine whether effects of poly(I:C) on natural cytotoxicity and tumor growth were mediated through production of interferon, anti-interferon antiserum (alpha-IF) was inoculated prior to poly(I:C) therapy. Tumor growth appeared to be uneffected by this maneuver even though poly(I:C)-induced enhancement of natural cytotoxicity was inhibited. Taken together, both macrophages and interferon may play a pivotal role in the immune response in both a stimulatory and a suppressive capacity. Additional study on such influences may be important if in vivo manipulation of these regulatory effects is to be accomplished successfully.

Modulation of stimulatory effects of poly(I:C) on natural cytotoxicity by anti-interferon.

The effect of polyinosinic . polycytidylic acid [poly(I:C)] on tumor inhibition in the context of natural cytotoxicity enhancement prompted further assessment of mechanisms underlying these effects. In vivo inoculations of poly(I:C) led to dose-dependent cytotoxicity enhancement in splenic lymphocytes and nonrecruited peritoneal exudate cells (monocytes). Although cytotoxicity of macrophages and lymphocytes together was less than that seen with lymphocytes alone, addition of indomethacin to these samples did not enhance cytotoxicity. In vivo inoculation of anti-interferon prior to poly(I:C) treatment prevented poly(I:C)-induced enhancement of natural cytotoxicity. Tumor growth was significantly inhibited by poly(I:C) treatment. Prior inoculation of anti-interferon antiserum partially prevented such tumor inhibition. Taken together, the tumor-inhibitory effect of poly(I:C) in this model may be mediated by interferon production and, at least in part, by interferon-induced enhancement of natural cytotoxicity.

Expression of the cellular immune response during tumor development in an animal model of bladder cancer.

Papillary and solid transitional cell cancers were induced in the bladders of Fischer rats by ingestion of the carcinogen N-[4-(5-nitro-furyl)-2-thiazolyl] formamide. At sequential regular intervals, purified splenic, peripheral blood, regional lymph node, and thymus lymphocytes were harvested from tumor-bearing and age-matched control animals and tested for cytotoxicity against a variety of cell lines. Lymphocytes were also tested in mixed lymphocyte cultures for blastogenesis. Tumor development was accompanied by decreased lymphocyte cytotoxicity against bladder tumor targets at each interval tested. This decrease was in addition to the age-related decrease in cytotoxicity expressed by lymphocytes from control animals. No corresponding differences were seen in lymphocyte subpopulations as determined by immunofluorescence or in lymphocyte blastogenesis. Suppressor cell activity in tumor-bearing animals could not be demonstrated, and no cause-effect relationship between cytotoxicity depression and tumor development in these experiments was apparent.

Specificity of cell membrane antigens in prostatic cancer.

Mice bearing a 3-methylcholanthrene-induced fibrosarcoma (MCAM-7) transplant in the right leg underwent surgical excision of the tumor and showed specific resistance to subsequent challenges with that identical tumor line. An in vivo response to tumor-specific antigens (MCAM-7 antigen) solubilized by hypertonic potassium chloride was measured by 24-hour footpad swelling response in mice immunized to the tumor from which the antigens were extracted. These observations suggested that the transplantable MCAM-7 fibrosarcoma could produce immunity toward the solubilized MCAM-7 tumor antigens and that this tumor immunity could be measured by footpad swelling response to injection of the solubilized antigens, an indication of cell-mediated immunity. The footpad swelling response was also minotored in relation to the extent of tumor growth. Mice received MCAM-7 tumor transplants by injection of 5 times 10-6 tumor cells and were tested for footpad swelling response at intervals following tumor transfer. A significant footpad response to injected MCAM-7 antigens was present 10 days following tumor transfer; at this time signs of tumor growth were only minimally detectable. The footpad swelling response to injected antigens disappeared by 28 days following initial tumor transfer; at this time the tumor diameters were in excess of 1.0 cm. Surgical removal of tumor at this point promptly restored footpad responses within 24 hours. Similar techniques have been applied to patients bearing adenocarcinoma of the prostate, where skin testing was substituted for the measurement of footpad swelling in animals. Seven patients with known prostatic carcinoma were given intradermal injections of soluble tumor antigens extracted from their own tumors. Three of the seven patients exhibited a cutaneous delayed type hypersensitivity response to the injected autologous tumor extracts. No positivereactions were observed in response to solubilized components of control tissues, including benign prostatic hyperplasia. The significance of the demonstrated concomitant immunity in these patients has not been resolved. However, these observations suggest that some patients bearing adenocarcinoma of the prostate can exhibit an immunologic response to specific antigens present in their own neoplasms.

Modification of lymphocyte responsiveness by hormone used in the treatment of urologic malignancies.

Recognition that the immune system may be important in regulating the clinical evolution of tumors and that corticosteroids suppress immune responses lends relevance to determining the concentrations in which sex hormones, currently used in the treatment of urologic malignancies exert immunosuppressive effects. We have investigated the effects of the sex hormones, diethylstilbestrol, diethylstilbestrol diphosphate, testosterone and progesterone, on in vitro lymphocyte blastogenesis as determined by 3H thymidine incorporation after stimulation with phytomitogens and alloantigens, and compared their effects to the effects of cortisol. Compared to cortisol these sex hormones are relatively weak suppressors of lymphocyte blastogenesis (cortisol 10(-7) M, progesterone 5 times 10(-6) M, testosterone 5 times 10(-5) M, diethylstilbestrol 5 times 10(-5) M and diethylstilbestrol diphosphate 10(-3) M) and probably are not significantly immunosuppressive in commonly used pharmacologic dosages. Similar results were observed with the T lymphocyte mitogens, phytohemagglutinin and concanavalin A, and in the combined T and B cell mitogen pokeweed. The fact that alloantigen-stimulated lymphocyte blastogenesis also was suppressed by diethylstilbestrol indicates that sex hormones exert their effects on the lymphocytes and not on the mitogens. Furthermore, sex hormones were not found to be cytotoxic to lymphocytes. It is postulated that the sex hormones tested act by suboptimal binding to glucocorticoid receptors in the lymphocytes and that the relative immunosuppressive potency of a given hormone is related to its affinity for the glucocorticoid receptor.

Castration effects on tumor specific immunity.

Using two immunogenic methylcholanthrene-induced fibrosarcomas in CD2F1 male mice, initial observations suggested that the rate of tumor growth might be enhanced by castration. For confirmation, tumor transplantation experiments using more than 500 mice were carried out in order to compare tumor specific transplantation immunity in castrate and in control male mice. Inbred mice bearing a 3-methylcholanthrene-induced fibrosarcoma transplant underwent surgical excision of the tumor; and specific resistance to subsequent challenges using varying doses of that tumor cell line were compared in castrate and in noncastrate groups of mice. Although castration influenced the rate of tumor growth, castration had no apparent effect on tumor specific immunoresistance. Mechanisms of host-tumor immunorelationships are discussed.

Calcium channel blockers inhibit cellular uptake of thymidine, uridine and leucine: the incorporation of these molecules into DNA, RNA and protein in the presence of calcium channel blockers is not a valid measure of lymphocyte activation.

An important role of transmembrane flux of calcium in lymphocyte activation has been previously demonstrated. Herein, we demonstrate that the calcium channel blockers verapamil and isradipine are able to inhibit in a concentration-dependent manner 3H-thymidine incorporation into DNA in phytohemagglutinin (PHA)-stimulated human peripheral blood mononuclear cells (PBMC). However, verapamil and isradipine diminish PHA-stimulated thymidine incorporation into DNA to the same extent whether they are added at the beginning of culture or 4 h prior to completion of a 72-h culture. Thus, 3H-thymidine incorporation into DNA in the presence of verapamil or isradipine is not a valid measure of mitogen-induced lymphocyte proliferation. Similarly, verapamil and isradipine also inhibit PHA-stimulated incorporation of 3H-leucine into protein and 3H-uridine into RNA whether the drugs are added at the beginning of culture or 4 h prior to completion of 24-h cultures. There is no intracellular accumulation of 3H-thymidine, 3H-leucine, or 3H-uridine into 10% trichloroacetic acid-soluble molecules during inhibition with verapamil or isradipine, suggesting that these drugs impair the cellular uptake of these substances rather than directly inhibiting their incorporation into DNA, protein, or RNA, respectively. Since previous reports documenting the inhibitory effects of calcium channel blockers on lymphocyte proliferation have utilized 3H-thymidine incorporation into DNA to measure proliferation, we have re-examined the antiproliferative effects of these drugs by determining their effect on PHA-stimulated cell cycle progression, employing cytofluorometric analysis of propidium iodide-stained cells. When added at the initiation of culture, both verapamil and isradipine inhibited in a concentration-dependent manner PHA-stimulated cell cycle progression.(ABSTRACT TRUNCATED AT 250 WORDS)

Identification of complement-receptor lymphocytes (B cells) in lymph nodes and tumor infiltrates.

To monitor the distribution of lymphocyte sub-populations infiltrating tumors, a simplified modification of standard techniques was developed to identify the complement-receptor B lymphocytes in frozen tissue sections. This technique is based on binding of antibody-coated bovine erythrocytes by B lymphocytes in the presence of sub-hemolytic concentrations of mouse complement. B lymphocytes can be identified in frozen tissue sections by adherent sensitized bovine erythrocytes. The specificity of the technique for B lymphocytes was established by 1) absence of spontaneous rosette formation with bovine erythrocytes by human T lymphocytes, 2) specific binding of bovine erythrocyte-antibody-complement complexes to B cell regions of normal lymph nodes and spleen and 3) binding of erythrocyte-antibody-complement to a high proportion of chronic lymphocytic leukemia (a B cell leukemia) lymphocytes. Five transitional carcinomas studied had regions of mononuclear infiltration that were virtually devoid of complement-receptor lymphocytes. The results suggest that lymphocytes infiltrating bladder carcinomas may be predominantly T lymphocytes.